Simple data manipulation is needed

[just4li]

This is an enhancement request.
In IoT Analytic, we have a use case to display the lab temperature with its Z-score. Z-score is in the range of (-4 to 4), the temperaure is in the range 16C to 23C.
In Zeppelin, we can easily do the following:
%sql
select index, temperature as Temperature, if ((zscore>3), zscore+17.5, 17.5) as Alert

It will overlay the Alert (ZScore) on top of Temperature so the chart is easy to understand.

In Brunel, this kind of data manipulation has to be done to the DataFrame before passing it to a Brunel plotting command. You have to have the new column ready. It is a design choice, but manipulating the existing data inside Brunel can give data scientist a lot of flexibility to discover the data trend.

[grahamwills]

It's always a tricky line to walk -- the degree to which you add analytic capability within a visualization system. With a background in machine learning and predictive analytics, I'm always tempted to add in smooth fit lines, nearest neighbor analysis and other cool techniques, so I'm strongly tempted by your suggestion.

However, I think we're going to pass on this one; at least for the next release. Since it is possible to do in the DataFrame, and, especially for big data / spark / Hadoop systems it is something you really should do in the DF, it gets a little lower priority than things that you cannot do at all.

I'm leaving this in the features list though -- other users have request multiple Y axes charts, and either that, or a combination of features, may make this application easier. So I'll assign it to a "future milestone" as a way of noting that we'll look at it later.

[grahamwills]

this might be tricky to get into 2.0 ... but will give it a go